Proceedings of the Twenty-Third International Joint Conference on Artificial Intelligence Human Action Recognition Using a Temporal Hierarchy of Covariance Descriptors on 3D Joint Locations

Subfields: Computer Vision and Pattern Recognition, 3D Vision & Pose

Citations in the corpus

• Space-Time Representation of People Based on 3D Skeletal Data: A Review (2016, methodology): “time dimension [121], [129], [190], [202], [208].”
• When Kernel Methods Meet Feature Learning: Log-Covariance Network for Action Recognition From Skeletal Data (2017, methodology): “The reason for this lies in the statistical computation of second order temporal momentum of Pa, the latter being very discriminative in recognizing human actions [16, 37, 3].”
• Modeling Temporal Dynamics and Spatial Configurations of Actions Using Two-Stream Recurrent Neural Networks (2017, background): “Joint based approaches consider the human skeleton as a set of points and use various positions based features such as joint positions [20, 31] and pairwise relative joint positions [46, 51] to characterize actions.”
• PKU-MMD: A Large Scale Benchmark for Continuous Multi-Modal Human Action Understanding (2017, background): “Those local features includes rotations and translations to represent geometric relationships of body parts in a Lie group [39, 40], or the covariance matrix to learn the co-occurrence of skeleton points [14].”
• A New Representation of Skeleton Sequences for 3D Action Recognition (2017, methodology): “Hand-crafted Features In [12], the covariance matrices of the trajectories of the joint positions are computed over hierarchical temporal levels to model the skeleton sequences.”
• Scene Flow to Action Map: A New Representation for RGB-D Based Action Recognition with Convolutional Neural Networks (2017, background): “Recognition of human actions from RGB-D data has generated renewed interest in the computer vision commu- ∗Corresponding author †To be appear in CVPR2017 nity due to the recent availability of easy-to-use and lowcost depth sensors (e.g. Microsoft Kinect TMsensor).”
• Deep Learning on Lie Groups for Skeleton-Based Action Recognition (2016, background): “Due to the development of depth sensors, 3D human activity analysis [27, 45, 23, 43, 41, 3, 42, 37, 44, 26, 35, 17] has attracted more interest than ever before.”
• Tensor Representations via Kernel Linearization for Action Recognition from 3D Skeletons (2016, background): “Keywords: Kernel descriptors, skeleton action recognition, higher-order tensors”
• Human Action Recognition by Representing 3D Skeletons as Points in a Lie Group (2014, methodology): “These approaches try to model the motion of either individual joints or combinations of joints using various features like joint positions [5] [8], joint orientations with respect to a fixed coordinate axis [20], pairwise relative joint positions [19] [22], etc.”
• WALKING WALKing walking: Action Recognition from Action Echoes (2017, methodology): “…representations or discriminative joint subsets from the skeletons to capture the correlation of body joints, and evaluate the similarity of different actions using a suitable metric [Devanne et al. , 2013; Hussein et al. , 2013; Vemulapalli et al. , 2014; Gong et al. , 2014; Zhang et al. , 2016].”
• Bregman Divergences for Infinite Dimensional Covariance Matrices (2014, background): “Such descriptors take the form of, e.g., region covariance matrices for pedestrian detection [24] and texture categorization [8], human joint covariances for activity recognition [10], and covariance matrices of the local Brownian motion of water molecules in diffusion tensor imaging (DTI) [20].”
• Deep Progressive Reinforcement Learning for Skeleton-Based Action Recognition (2018, methodology): “We adopted the leave-oneout cross-validation protocol to evaluate our method on this dataset [45].”
• Real-Time Multi-scale Action Detection from 3D Skeleton Data (2015, methodology): “[14] provided ground truth annotations for the start and end frames of each action in the MSRC-12 dataset.”
• Skeleton-Based Action Recognition with Spatial Reasoning and Temporal Stack Learning (2018, methodology): “[10] use the covariance matrix for skeleton joint locations over time as a discriminative descriptor for a sequence.”
• Efficient Temporal Sequence Comparison and Classification Using Gram Matrix Embeddings on a Riemannian Manifold (2016, background): “[17] built a temporal hierarchy of covariance descriptors on 3D joints to classify human actions.”